Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{204619,
        author = {H B Pramod and D Suresha},
        title = {Hybrid Deep Ensemble Learning Framework with Explainable AI for Adaptive Optimization of Underwater Wireless Sensor Networks},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {3533-3541},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=204619},
        abstract = {In the context of marine ecosystems monitoring, oceanographic research and inspection of subsea infrastructure, Underwater Wireless Sensor Networks (UWSNs) are key enablers. This paper presents a Hybrid Deep Ensemble Learning Framework (HDELF) using Bidirectional LSTM (BiLSTM), Attention-CNN, TabNet, Gradient Boosting Models (CatBoost, XGBoost, LightGBM) and classical classifiers for holistic performance prediction of the UWSN. An Explainable AI module with SHAP gives interpretable feature attribution, and Bayesian hyperparameter optimisation and federated learning adds to the practicality. The Attention-CNN is trained on 18 features and 150,000 samples, and is able to reach a 94.3% accuracy (ROC-AUC score 0.96), with the three most predictive features being identified as SNR, battery level, and water depth based on SHAP.},
        keywords = {Attention Mechanism; BiLSTM; CatBoost; Explainable AI; Federated Learning; SHAP; TabNet; Underwater Wireless Sensor Networks; XGBoost},
        month = {June},
        }